Skip to content
PaperFren

Concept

Sleep Architecture

7 studies3 discoveriesEvidence last moved Sep 20, 2026

Sleep architecture is the organisation of a night's sleep into stages and cycles — how much slow-wave and REM sleep occurs, in what order, and how often it is interrupted. It is what polysomnography measures and what a report of hours slept does not.

Studies that ask whether sleep helps cognition usually manipulate duration, because duration is easy to manipulate. The associations in this library sit with composition and continuity instead, which is why 'get eight hours' and 'sleep well' are not the same recommendation and do not have the same evidence behind them.

Studies

7

Findings

5

6 supporting · 0 challenging · 0 qualifying citations

Open tensions

1

Latest change

Concept page published

Sleep Architecture

Currently

What we know

  1. The association sits with how the night was structured, not simply how long it lasted.
  2. Fragmentation accompanied accelerated forgetting even where short-delay memory looked normal.
  3. Total and partial deprivation are different manipulations with different effect sizes.
  4. Which measure you take determines which countermeasure appears to work.
  5. A self-regulation failure at bedtime predicted the same outcomes usually attributed to sleep need.

Largest unresolved question

Whether sleep loss impairs cognition depends on the population and the outcome, and these studies do not share either. Total deprivation degraded accuracy in young adults across three memory tasks, while in older adults a night of total deprivation left working-memory accuracy, speed and the positivity bias unchanged on a brief task.

Common misconceptions

  • Hours of sleep is the measure that matters for memory.

    In the polysomnography study the association ran through slow-wave sleep percentage (r = 0.61 with combined recall), and in the ageing study accelerated forgetting accompanied more frequent night-time awakenings. Both are composition and continuity, not duration.

  • A night without sleep impairs everything you can measure.

    After total sleep deprivation, older adults' working-memory accuracy and speed and their positivity bias were unchanged on a roughly 12-minute task. The authors note this does not extend to longer or more demanding challenges.

  • Caffeine and a nap are interchangeable ways to recover from sleep loss.

    They improved different measures — caffeine simple reaction time, the nap two-choice reaction time and, after normal sleep, mood — and combining them was not simply additive.

Related